A/B Testing Website Load Time (Per-User) Calculator
Compare control and variant page load times to estimate the time saved or added for each user in an A/B test.
Overview
Use this A/B testing website load time calculator to compare the average load time of a control page with a variant. Enter the test traffic split, user count, and average page loads per user to see the time difference per user and the cumulative impact for variant visitors.
How it works
The calculator subtracts the variant load time from the control load time. It then multiplies that difference by the average page loads per user to estimate the time saved or added per user. The percentage change is measured against the control load time. Total time saved is the per-user difference multiplied by the estimated number of users assigned to the variant. Positive results mean the variant is faster; negative results mean it is slower.
How to use this calculator
- 1Enter the average load time for the control version.
- 2Enter the average load time for the variant version.
- 3Add the total number of users included in the test.
- 4Set the percentage of traffic assigned to the variant.
- 5Enter the average number of measured page loads per user and review the results.
Example Calculation
Control load time
3
Variant load time
2.4
Total test users
10000
Variant traffic share
50%
Average page loads per user
1
Time Saved Per User
0.60 seconds
With a control load time of 3.0 seconds and a variant load time of 2.4 seconds, the variant saves 0.60 seconds per user. At a 50% split, 5,000 variant users save about 3,000 seconds in total.
Frequently asked questions
What does this A/B testing load time calculator measure?
It estimates how much faster or slower a variant is per user compared with a control, plus the cumulative time change for users assigned to the variant.
What does a negative time saved per user mean?
A negative result means the variant is slower than the control by that amount for each user across the entered number of page loads.
Why should I enter page loads per user?
A load-time difference can occur more than once during a visit. This field lets you estimate the total impact when each user loads the measured page multiple times.
Does a faster page automatically improve conversions?
No. Faster performance may improve user experience, but the calculator does not determine whether any observed conversion or engagement change is statistically reliable.
Which load-time metric should I use?
Use one consistent metric for both versions, such as a page-load measurement or a chosen performance metric. Do not compare different metrics between control and variant.
Why is the traffic split included?
The traffic split estimates how many test users saw the variant, which is needed to calculate the cumulative time saved or added across variant users.
Explore Related Calculators
Assumptions and warnings
Assumptions
- Control and variant load times are measured using the same page, load-time metric, devices, and reporting period.
- Each variant user is assumed to experience the entered average number of measured page loads.
- The traffic split is assumed to be representative of the actual test allocation.
- The calculation measures time differences only; it does not estimate conversion, revenue, retention, or statistical significance.
- Results are estimates based on average values and may not reflect individual user experiences.
Warnings
- A faster measured load time does not by itself prove that the variant improves business outcomes.
- Check sample size, test duration, measurement consistency, and statistical significance before acting on A/B test results.